<bitcoin-git>
[bitcoin] maflcko closed pull request #32006: docs: remove passing CI section in guidelines for PR merging as it's common sense (master...fix/typos) https://github.com/bitcoin/bitcoin/pull/32006
<bitcoin-git>
[bitcoin] maflcko reopened pull request #32006: docs: remove passing CI section in guidelines for PR merging as it's common sense (master...fix/typos) https://github.com/bitcoin/bitcoin/pull/32006
eugenesiegel has quit [Quit: Client closed]
eugenesiegel has joined #bitcoin-core-dev
bugs_ has joined #bitcoin-core-dev
diego has left #bitcoin-core-dev [#bitcoin-core-dev]
dviola has joined #bitcoin-core-dev
<darosior>
TheCharlatan: motivation seems worthwile to me, but how does changing ConnectBlock() to take spent coins through undo data instead of through a map make things any easier for clients without a full UTxO set at hand?
eugenesiegel has quit [Quit: Client closed]
eugenesiegel has joined #bitcoin-core-dev
eugenesiegel has quit [Quit: Client closed]
Talkless has joined #bitcoin-core-dev
aleggg has quit [Remote host closed the connection]
<jeremyrubin>
you can very likely safely change it to fail if the vtx.size() is greater than something like 25,000
<sipa>
minimum transaction vsize is 60, i think? so the limit should be 16665
<instagibbs>
sounds right ^
<darosior>
Yeah it's basically assuming a minimum size of 1 byte for txs. Whereas it's larger, so probably safe to decrease. But i agree with instagibbs
<jeremyrubin>
I think you can do smaller than 60 iirc
<instagibbs>
darosior ah, makes sense
* darosior
looks up SE, Murch probably has answered the minimum valid size there
<eugenesiegel>
lightlike: I was thinking about MAX_SIZE, not MAX_PROTOCOL_MESSAGE_LENGTH woops
twistedline has quit [Ping timeout: 265 seconds]
<eugenesiegel>
instagibbs: There is also a check that the BlockTxCount doesn't overflow uint16 so I think the maximum number of compact block txns is 65535
twistedline_ has joined #bitcoin-core-dev
NodeDiver has quit [Quit: Client closed]
<TheCharlatan>
darosior, you could just insert the utxos into the coins cache's map too, but that seems awful. If such clients do full validation they usually require the utxos to be sent alongside a block. They just don't store them permanently.
eugenesiegel has quit [Ping timeout: 240 seconds]
eugenesiegel has joined #bitcoin-core-dev
<darosior>
I see
cotsuka has quit [Remote host closed the connection]
cotsuka has joined #bitcoin-core-dev
eugenesiegel has quit [Ping timeout: 240 seconds]
eugenesiegel has joined #bitcoin-core-dev
<achow101>
instagibbs: sipa: I don't think you can have both the prevout's scriptpubkey and the input scriptsig be empty though?
<achow101>
i believe my stackexchange answer is saying the prevout is an OP_TRUE, which allows the scriptSig to be empty
<instagibbs>
input scriptSig empty, new output empty, you can do that and that's 60
Christoph_ has joined #bitcoin-core-dev
<achow101>
yeah, that's what my stackexchange answer says
<sipa>
achow101: sure, but the prevout's scriptpubkey doesn't need to be in the same block/transaction
<achow101>
right, but I didn't say that :)
<achow101>
oh antoine just edited that
<achow101>
nevermind
<instagibbs>
hah
<achow101>
well, I still didn't say that the spending tx has its prevout in the same block
Christoph_ has quit [Quit: Christoph_]
Talkless has quit [Quit: Konversation terminated!]